Why microcap stocks trade thin, and how to measure it

01Two numbers that matter

The first is turnover: median daily dollar volume divided by market value, usually shown in basis points. It lets you compare a $60 million company with a $600 million one on the same scale, which raw volume cannot do.

The second is days to turn the register: shares outstanding divided by median daily share volume. It says how many trading days it would take for every share to change hands once. A fund reads it as a rough answer to one question: if we bought a meaningful position, how long would it take to get out?

Use the median, never the average. One day with a press release or an index change can move the average for months. The median is the kind of day a fund actually trades in.

02What the range looks like

Across the US-listed companies Krypton AI measured in September 2026, days to turn the register ran from 29 trading days to more than 2,400. At the slow end, the whole register changes hands about once every ten years.

Among companies under $500 million, the ones an analyst covers trade a median of $1.05 million a day, and the ones nobody covers trade $157,000. Liquidity and coverage move together because they feed each other: an analyst brings institutional readers, and institutions need a stock they can trade before an analyst will write for them.

03Why it keeps funds away

A fund that wants 2% of a $100 million company needs $2 million of stock. At $150,000 a day it can only buy a fraction of that on any one day without moving the price, and it has to believe it can sell just as slowly.

Quantitative funds trade on models and screens. They buy what they can trade, so they rarely solve thin trading for a company. Index funds hold what the index holds and trade when it rebalances. The holders that change a company's liquidity are the ones that pick stocks and hold them.

04Getting the share count right

Days to turn is only as good as the share count under it, and the count is easy to get wrong. In our own work, the share count a company reports in its XBRL data was wrong in three different ways across eight companies in two days.

• The tag can simply stop being updated. For one company the latest value was from 2012, and it still looked like a valid number.
• The date can be nonsense. One filer's latest value was dated 2033.
• Companies with two share classes often report only one. Two we checked were 15% and 27% light as a result.

05The one-line check

Multiply the share count by the share price and compare the result with the company's market value. If the ratio is outside roughly 0.85 to 1.18, the count is wrong. When it fails, go to the cover page of the latest 10-K or 10-Q and add up every class printed there.

It matters more than it sounds. On one company, a stale count gave 178 days to turn and the real count gave 72: the difference between a stock a fund will look at and one it will not.

06Float is part of the answer

Low volume is sometimes a float question before it is an attention question. If directors, officers and one or two strategic holders own a large block, less stock is left to trade.

Institutional stock is float. Index funds and active funds both sell. The defensible measure is shares outstanding less the proxy's director and officer group, less any strategic holder who does not trade.

When the float is small for a good reason, say so. A company that publishes its float and explains it gives a fund one less thing to work out alone.

07What moves it

Volume follows people who have a reason to own the stock. That means more funds that know the story, analysts who compare it with peers, and individual investors who can find something to read about it. None of it is fast, and a company can track all of it: turnover against peers, days to turn, and how many funds that pick stocks own it each quarter.